Development of SRF Cavity Technology Four decades of Progress with Prof. Y. Kojima s Pioneering Work. Peter Kneisel Jefferson Lab

Transcription

1 Development of SRF Cavity Technology Four decades of Progress with Prof. Y. Kojima s Pioneering Work Peter Kneisel Jefferson Lab

2 Contents

3

4 Our Friendship (1) Prof. Kojima Yuzo- was my friend for 35 years He was not only my friend, but also a friend of our family He visited us often in Karlsruhe, Ithaca and Williamsburg We tried to pursuade him to bring his wife on some of his visitsinstead one time he came with his daughter Naoko One time he came with Kenji, who then decided that he wanted to work for some time at Jlab He invited me to stay at KEK for two months in 1980 He took me to a dance recital of his daughter Naoko We had many wonderful hours together: bowling in Karlsruhe, bicycling in Ithaca, getting snowed-in in Ithaca, celebrating his birthday on a boat on Cayuga Lake, drinking everywhere

5 Our Friendship(2) When he stayed in our house, he cooked for us. Chicken a la Yuzo is one recipe in our cookbook, which we like very much

6 Our Friendship (3) When I was a guest at KEK for 2 months in 1980, he showed my many things: We rode the Shinkansen together to Kobe We hiked Mount Fuji half way We visited Tohoku University and on the way back, stopped in a riokan He introduced me to Sushi and Tekka Don He showed me and taught me the importance of personal relationships He arranged my visit to Tokyo University by train, giving me meticulous written instructions in Japanese in case I would get lost He was a wonderful host

7 Our Friendship(4) I met Yuzo the first time in 1973 at Stanford University/HEPL, where he was on leave of absence to learn the technology for sc cavities and accelerators. I came to HEPL as a post doc to introduce electropolishing to HEPL In those days, HEPL was the center and frontier of SRF Technology The HEPL colleagues had demonstrated in 1972 in an X-band cavity, that very high gradients and Q-values could be obtained with niobium as the superconducting material Based on this extraordinary success HEPL proceeded with the design and construction of the SCA ( superconducting accelerator), operating at 1300 MHz and 2K However, this effort ran into major difficulties caused by Multipacting

8 HEPL X-band cavities Weissman, J. P. Turneaure; A Nb TM 010 Mode Cavity with High Electric Field and Q 0, Appl.Phys.Lett 13, 390 (1968) J.P. Turneaure, N.T. Viet, Superconducting Nb TM 010 Mode Electron-beam Welded Cavities, Appl.Phys.Lett 16, 333 (1970)

9 1. Introduction Cavities Superconductivity RF Acceleration 1961: Bill Fairbank (Stanford Univ.) presented the first proposal for a superconducting accelerator. 1964: Bill Fairbank, Alan Schwettman and Perry Wilson (Stanford University) First acceleration of electrons with sc lead cavity. 1967: John Turneaure (Stanford University) Epeak =70 MV/m and Q~10 10 in 8.5 GHz cavity!! : Mike McAshan, Alan Schwettman, Todd Smith, John Turneaure and Perry Wilson (Stanford University) Development and Construction of the Superconducting Accelerator SCA. and today Seminar I+II: Superconducting elliptical cavities and HOMs Jacek Sekutowicz, Beijing, Mianyang, November, /115

10 SRF Worldwide Courtesy of JLab T

11 HEPL At HEPL, Prof. Kojima worked with a young graduate student Claude Lyneis on S- and L-band cavities and learnt from him the technology of chemical polishing, high temperature heat treatment and testing. He was co-author of a paper, presented in 1973 at the PAC in San Francisco: C. Lyneis, Y. Kojima, J.P. Turneaure, N.T. Viet Electron Loading in L- and S- Band Superconducting Niobium Cavities ; Proc. PAC 1973, p. 101 The main problems at the time was multipacting and field emission

12 HEPL Here are some of the memories of Claude Lyneis about Prof. Kojima s visit: Yuzo had been a full professor and a boss ( at Tohoku University, responsible for electron linac development) prior to coming to Stanford, but he was very happy to work with his hands in the laboratory I guess Yuzo and I were measuring the Bremsstrahlung from S-Band and maybe L-Band. I do remember the excitation of the higher order modes in the S-band cavities. There are some impressive radiation levels quoted in the paper. That might shut us down these days. I was working on x-band cavities R, T and delta X, T measurements for my thesis and Yuzo took on much of the S-band measurement work. He was a great person to work with, very humble for a professor helping a graduate student

13 Preparation for Tristan Upgrade (1) After return from HEPL, Prof. Kojima established a small SRF group The group studied fabrication and treatment procedures on C-band ( 6 GHz) single and multi-cell cavities In single cell cavities Q values of 2 x10e10 and Eacc = 10 MV/m were achieved In an acceleration test with a 9-cell cavity Eacc = 3 MV/m were reached T.Furuya, K. Hosoyama,T.Kato, Y.Kojima and O. Konno Proc Linear Acc. Conf, Montauk(1979), 194 At the end of 1979 efforts focused on the possibility of increasing the energy of TRISTAN from 30 GeV to >35 GeV by adding to the normal conducting cavities 500 MHz superconducting niobium cavities

14 Preparation for Tristan Upgrade (2) In my view, Prof. Kojima directed the work in preparation for for TRISTAN to the following areas: Establish a capable and powerful R&D and production group Develop the proper surface treatment procedures: electropolishing initially vertical, but subsequently developed by Saito-san to continuous horizontal EP the method used nowadays for high performance cavities for the X-FEL and ILC Improvement of thermal conductivity by post-purification with Ti Develop a robust and powerful input coupler these couplers, with some modifications, have been adopted by SNS and the KEKB projects Develop with industry multi-cell cavities, industrial surface preparation, (Nomura Plating Comp) cryostats, cryo-modules and assembly procedures It was no accident that during my 2 months visit he took me to Tohoku University, Nomura Plating and MHI in Kobe he very much believed in personal relationships with his industrial partners. Develop an appropriate cryogenic system

15 Prototype 500 MHz Cavity (1) A spherical shape was chosen to reduce or eliminate multipacting and an elliptical transition at the beam pipe to minimize peak electric fields This cavity was fabricated by MHI the electropolishing of the half cells took place at Nomura Plating and I had the good fortune to witness a large part of the fabrication process Prior to the prototpye development at KEK, Prof. Kojima had spent in 1979 a few months at the Kernforschungszentrum Karlsruhe (KfK) and participated in the development of storage ring cavities for DORIS At KfK he worked together with Noguchi-san a visitor at the time from Tokyo University, following the visits from Yoshida-san and Yoshioka-san on measurements and surface preparation both electropolishing and bcp. This DORIS cavity was qualified for insertion in the storage ring by careful measurements of the electron loading, heating, X-ray distribution and trajectory calculations. The result of this work was a publication in Nuclear Instr. & Methods: S.Noguchi, Y.Kojima und J.Halbritter, Nuclear Instruments and Methods 179,205(1981)

16 Prototype 500 MHz Cavity (2)

17 Prototype 508 MHz Cavity T.Furuya, S.Hiromatsu, T.Nakasato, T.Kato, P. Kneisel, Y.Kojima and T.Takagi; First results on a 500 MHz Superconducting Test Cavity for TRISTAN ; PAC 1981 The cavity was a success: it reached the design gradient for TRISTAN of Eacc = 3 MV/m at 4.2K with a Q-value of Q = 2.8 x 10e9 and very moderate electron loading

18 TRISTAN Upgrade(3) In the summer of 1988 sixteen 5-cell cavities were installed in the TRISTAN tunnel The performance of these 16 cavities and the results of early beam operation were reported at PAC 1989: Y. Kojima, K. Akai, M. Arinaga, K. Asano, E. Ezura, T. Furuya, K. Hara, K. Hosoyama, A. Kabe, E. Kako, K. Kubo, S. Kurokawa, S. Mitsunobu, H. Nakai, T. Nakazato, S. Noguchi, T. Ogitsu, K. Saito, Y. Sakamoto, T. Shishido, T. Suzuki, T. Tajima, T. Takashima ; Upgrading of TRISTAN by Superconducting RF System, PAC 1989, Chicago, p.1789 The beam energy of TRISTAN was upgraded from 28.5 GeV to 30.7 GeV The second set of 16 cavities were installed at the beginning of August 1989 and all the participants of the 4 th SRF workshop at KEK in August 1989 had a chance to visit the tunnel and admire the installations (including myself).i n the fall of 1989, a beam energy of 32 GeV was achieved.

19 Tristan Upgrade (4) The majority of the 32 cavities reached a gradient of 10MV/m in the vertical test; the assembly into pairs with input couplers, HOM couplers and tuners was done in a class 100 clean room; some assembly steps needed to be done in a less clean environment; this might have resulted in the observed 30% degradation The upgrade of TRISTAN was the first large scale successful demonstration of SRF technology in an accelerator and was truly a pioneering effort due to a visionary leadership by Prof. Kojima and a dedicated and immensely competent group of collaborators at KEK. A few years later ( ~ 1993) Jlab s CEBAF with 338 five-cell cavities, operating at 1497 MHz, 5 MV/m and 2K came on-line. Saito-san helped for 2 years in the efforts to qualify cavities. The storage ring HERA in Hamburg was commisssioned around the same time ( 500 MHz, 5-cell cavities, Eacc ~ 5 MV/m)

20 SRF Technology Developments(1) While the low gradient accelerator projects TRISTAN, CEBAF and HERA were implemented, many efforts went into fighting the major enemies of inferior cavity performance and to improve the technology: cavity shape(multipacting), insufficient material removal, defects including fabrication inadequacies such as e.g electron beam welds and contamination Even though big progresses have been made shifting cavity performances to higher gradient levels, these difficulties still exist, Courtesy of K.Saito,F.Furuta

21 SRF Technology Developments(2) So, what were the technology improvement? Understanding and eliminating Multipacting Understand and find cures for Q-disease Thermally stabilize the material against defects ( niobium with higher thermal conductivity, post-purification,eddy current scanning,thermal model calculations) Diagnostics (temperature mapping, X-ray mapping, guided repair ) Reduce contamination to reduce field emission ( high pressure water rinsing, clean room assembly, high peak power processing..) Improve assembly procedures Develop improved surface treatment procedures (EP, CBP, Large grain) Investigate, understand and eliminate the high field Q-drop Optimize cavity shapes ( peak field ratios, multipacting behaviour) to reduce magnetic field levels ( LL, Ichiro, re-entrant)

22 Multipacting/Resonant Electron Loading At a certain combination of electric and magnetic field strength at the outer wall of a cavity, electrons can be resonantly with the rf cycles moved into the cavity volume and accelerated back to the surface. If the secondary electron emission coefficient of the niobium is >1 at the impact energy of the electrons, an avalanch process gets started, all the rf energy in the cavity is absorbed in this process and a severe limitation in gradient can occur It was Claude Lyneis at HEPL, who first simulated this process with a tracking program and could explain the limitations observed in the HEPL L- band cavities He also showed that the elliptical cavity shape of the Genoa group [R.Parodi et al, 1979] lacked the combination of fields to generate multipacting Other calculations followed at CERN (J.Tueckmantel,) Cornell (H.Padamsee et al,) Uni Wuppertal (G. Mueller et al) This knowledge is since used to design multipacting-free cavities

23 Multipacting: C.Lyneis, SRF 1980 HEPL Cavity, 3 rd order MP at Eacc ~ 2.15 MV/m

24 Defects, Cleanliness,Surface Treatment With the elimination of MP, higher fields could be obtained and new barriers had to be overcome Improved thermal conductivity of the niobium both as manufactured and improved by post-purification stabilized defects; eddy current scanning (DESY) weeded out defective niobium sheets Temperature mapping helped to identify problems such as FE emitters, quench areas and gave a picture of a real cavity surface High pressure ultrapure water rinsing and clean room assembly attacked the contamination problems encountered in these cavities Improved surface treatments electropolishing and baking made it possible to reach gradients close to the fundamental material limits of niobium ( critical magnetic field)

25 Post purification of Nb [W.Singer, 2003] Thermal conductivity of samples from the niobium sheets used in the TESLA cavities: before and after the 1400 ºC heat treatment (RRR = 270 and RRR = 500 respectively) Eacc, MV/m Cavity post purification (solid state gettering) March 18, 2005 RRR quench pow er limit Eacc versus RRR of TTF cavities ERL 2005, Jefferson Lab The heat treatment also homogenize the Nb ( reduction of magnetic flux pinning centers shown by magnetization measurement)

26 Scanning of Niobium Sheets Successfully developed at DESY to pre-screen Nb Sheets for defects: eddy current, resolution ~ 100 µm squid, resolution < 50 µm (W.Singer, X.Singer) March 18, 2005 ERL 2005, Jefferson Lab

27 A Breakthrough At the 8 th SRF workshop in Abano Terme, Italy, Kenji Saito presented a paper for the KEK group, which had a large impact in the future direction of SRF technology: K. Saito,H.Inoue,E.Kako,T.Fujino,S.Noguchi,M.Ono,T.Shishido Superiority of Electropolishing over Chemical Polishing on High Gradients Systematic studies of varying amounts if chemical polishing on cavities, followed by electropolishing and /or barrel polishing clear ly showed that the Q-drop above ~ 25 MV/m ( European Headache ) could be overcome by electropolishing and gradients of > 40 MV/m were obtainable Electropolishing became the favorite surface preparation technique for high gradients

28 March 18, 2005 ERL 2005, Jefferson Lab

29 March 18, 2005 ERL 2005, Jefferson Lab

30 EP- Systems KEK/Nomura Plating DESY JLab INFN Cornell Lutz Lilje DESY -MPY March 18, 2005 ERL 2005, Jefferson Lab

31 High Pressure Water Rinsing Universally used as last step in surface preparation Water: ultrapure, resistivity > 18 MΩcm Pressure: ~ 100 bar ( 1200 psi) Nozzle configuration: varying, SS or sapphire Scanning : single or multiple sweeps, continuous rotation + up/down Add. HPR after attachment of auxiliary components March 18, 2005 ERL 2005, Jefferson Lab

32 High Pressure Rinse Systems DESY-System KEK-System Jlab HPR Cabinet March 18, 2005 ERL 2005, Jefferson Lab

33 Centrifugal Barrel Polishing(CBP)(1) Barrel Polishing ( tumbling ) developed at KEK for smoothening of surfaces/welds plastic stones, water + abrasive Process very slow, by adding motion, removal rate increased 10fold: ~ 44 µm in 8 hrs During the process, hydrogen is dissolved in the niobium( Q-disease ) and needs to be removed by furnace treatment Hydrogen-free CBP accomplished by using a different (hydrogen-free) agent:fc-77 (C8F18,C8F16 O) [T.Higuchi,K. Saito SRF 2003] March 18, 2005 ERL 2005, Jefferson Lab

34 Centrifugal Barrel Polishing(2) [T.Higuchi, K. Saito, SRF 2003 ] March 18, 2005 ERL 2005, Jefferson Lab

35 High Temperature Heat Treatment UHV Heat Treatment of Niobium used since the beginning of times ; nowadays : Hydrogen degassing: 600C for 10 hrs at Jlab 750 C for 3 hrs at KEK Annealing: 800 C, several hrs Post- Purification: 1200C to 1400C in presence of a solid state getter, e.g.ti Improvement of RRR Loss of mechanical properties grain growth March 18, 2005 ERL 2005, Jefferson Lab

36 3. New Shapes: Pros and Cons KEK single-cell tests in September 2005!!!!!!! LL single 1st cavity 15th, EP(30)+HPR+Bake Qo Courtesy K. Saito Qo 2K Qo 1.68K Quench 46.5MV/m 1.97K 1.68K E peak = 86.5 MV/m B peak = mt Eacc [MV/m] Seminar III: SRF for ILC Jacek Sekutowicz, Beijing, Mianyang, November, /88

37 TESLA After the successful demonstration of SRF technology in TRISTAN and the construction of CEBAF and HERA, Prof. B. Wiik, the DESY director, proposed a linear collider with unprecedented gradient goals for the cavities, based on the R&D achievements in the 80 s 11 year s of R&D after the first TESLA workshop at Cornell in 1990, lead to a technical design report presented to the German Government in 2001 e - gun, preacceleration for the e - damping auxiliary positron ring source Target for e + production e + damping ring e - source and acceleration 5 GeV e - main linac undulat or 33 km IPs e + main linac

38 2. TESLA Cavities and Auxiliaries as ILC Baseline Design The cavity was designed in 1992 (A. Mosnier, D. Proch and J.Sekutowicz). End-cell 1 7 identical inner cells End-cell 2 TTF 9-cells; Contour of E field f π [MHz] f π -1 [MHz] R/Q [Ω] 1012 G [Ω] 271 Active length [mm] 1038 Seminar III: SRF for ILC Jacek Sekutowicz, Beijing, Mianyang, November, /88

39 2. TESLA Cavities and Auxiliaries as ILC Baseline Design The inner cell geometry was optimize with respect to: low E peak /E acc k cc. f π [MHz] r iris [mm] 35 k cc [%] 1.9 E peak /E acc B peak /E acc [mt/(mv/m)] 4.15 R/Q [Ω] G [Ω] 271 and coupling At that time (1992) the field emission phenomenon and field flatness were of concern, no one was thinking about reaching the magnetic limit. R/Q*G [Ω*Ω] Seminar III: SRF for ILC Jacek Sekutowicz, Beijing, Mianyang, November, /88 Inner cell; Contour of E field

40 3. New Shapes: Pros and Cons Kenji Saito (KEK) proposed (PAC2003, TESLA Meeting 2003) to re-evaluate our criteria for the cavity design. Kenji argues that: The field emission is not a hard limit in the performance of sc cavities if the surface preparation is done in the right way. Unlikely this, magnetic flux on the wall limits performance of a sc cavity (Qo decreases or/and quench). Hard limit ~180 mt for Nb. B peak / E acc should be low 1. Cavities may operate at higher gradients. 2. Cavities may operate at lower cryogenic load. Seminar III: SRF for ILC Jacek Sekutowicz, Beijing, Mianyang, November, /88

41 3. New Shapes: Pros and Cons J.Sekutowicz, Lectures in China We know how to reduce B peak / E acc ) : more volume in equator region and smaller iris. TTF LL RE / r iris [mm] 35 k cc [%] 1.9 E peak /E acc B peak /E acc [mt/(mv/m)] 4.15 R/Q G R/Q*G [Ω] [Ω] [Ω*Ω] Seminar III: SRF for ILC Jacek Sekutowicz, Beijing, Mianyang, November, /

42 3. New Shapes: Pros and Cons KEK single-cell tests in September 2005!!!!!!! Qo Re-entrant 11th T = 1.8 K 2005/09/07 Eacc = MV/m Q0 = 6.88e9 Courtesy K. Saito Just adding LiHe 2K 1.8K T = 2.0 K 48 MV/m E peak = MV/m B peak = mt!? MV/m runout LiHe during proc Eacc [MV/m] Seminar III: SRF for ILC Jacek Sekutowicz, Beijing, Mianyang, November, /88

43 From TESLA to XFEL and ILC In 2003 the International Technology Recommendation Panel (ITRP) recommended the use of sc cavities for the next linear collider The original TESLA design morphed into 2 designs: X FEL and ILC The more ambitious of these is the ILC with a design goal of Eacc = 35 MV/m in vertical tests and Eacc~ 31.5 MV/m in a 8x9 cell cavity cryomodule An R&D program was established ( S0 for vertical tests and S1 for modules) to achieve these goals Even though on several occasions such gradients have been achieved, reproducibility is the major problem to be solved The GDE meeting next week here at KEK will have re-baselining on its agenda Alternatives to the baseline such as use of a different cavity shape (LL,Ichiro) or different material ( large grain ) are being pursued on a modest level

44 Preliminary RF statistic of 6th cavity fabrication XFEL Spec. Eacc=23,6 MV/m Accel-Zanon cavities Eacc,max (MV/m) fe fe fe fe 5 0 AC 116 AC 121 AC129 AC119 AC 123 AC 128 Z 133 Z 138 Z 135 Z 140 AC 120 AC 125 AC 122 AC 126 Z 131 Z 130 Z 144 Z 132 Z 139 Z143 Number of cavity AC BCP Flash E acc =30,2 +/- 4,9 Z BCP Flash E acc =24,9 +/- 3,8 - Max gradient, FE marked, if starts below 20 MV/m - With He-vessel - Without HOM pick up AC EP E acc =29,3+/- 9,7 Z EP E acc =24,9+/-4,4

45 Jlab/ILC Cavities: preliminary

46 Our Friendship Before I finish, I want to return to our friendship and tell you that even after Prof. Kojima had retired from KEK, we used every possible occasion to meet each other, be it at my visits to KEK or at joint participation in SRF workshops (2001 in Japan and 2003 in Germany) Yuzo was always interested in keeping up with SRF technology developments worldwide. Our last get-together was in 2006, when I presented a lecture at the ILC school in Sokendai. Prof. Kojima came by bus to meet me at the school, we then drove back to his house, where he proudly showed me the dance studio, he had established for Naoko, and then we went to dinner in Hayama

47

48 Final Remarks Prof. Kojima s leadership during the TRISTAN project has left deep marks in SRF cavity development and many presently pursued large scale projects benefit from his vision and persistance: electropolishing, high power couplers and industrialization efforts to name a few In his characteristically quiet style not looking for the limelight and for publicity - he has been a great role model for his colleagues working with him on the TRISTAN project and I believe that he has displayed the character traits, which are essential for any leader of a large scale project With his hands-on approach he has inspired his TRISTAN colleagues to do an outstanding job and to show the road to the future after all, TRISTAN was the first large scale, successful application of SRF technology

49 Final Remarks I believe and hope,that in these short memory times his pioneering contributions to SRF technology will not be forgotten in the SRF world I am absolutely certain, that in Japan his legacy will be highly guarded As far as I am concerned, I was so privileged to have Yuzo as a longtime friend and as long as I live, he will be in my heart.